Carrasco-Pancorbo 2005-review.pdf

8/10/2019 Carrasco-Pancorbo 2005-review.pdf

1/22

J. Sep. Sci. 2005, 28, 837858 www.jss-journal.de i 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

R e v i e w

Analytical determination of polyphenols in olive oils837

Alegria Carrasco-Pancorbo1

Lorenzo Cerretani2

Alessandra Bendini2

Antonio Segura-Carretero1

Tullia Gallina-Toschi2

Alberto Fernndez-Gutirrez1

1Department of AnalyticalChemistry, Faculty of Sciences,University of Granada,C/Fuentenueva s/n, E-18071Granada, Spain

2Department of Food Science,University of Bologna, P.zzaGoidanich 60, 47023 Cesena(FC), Italy

Analytical determination of polyphenols in olive oils

The increasing popularity of olive oil is mainly attributed to its high content of oleic

acid, which may affect the plasma lipid/lipoprotein profiles, and its richness in pheno-

lic compounds, which act as natural antioxidants and may contribute to the prevention

of human disease. An overview of analytical methods for the measurement of poly-

phenols in olive oil is presented. In principle, the analytical procedure for the determi-nation of individual phenolic compounds in virgin olive oil involves three basic steps:

extraction from the oil sample, analytical separation, and quantification. A great num-

ber of procedures for the isolation of the polar phenolic fraction of virgin olive oil, utiliz-

ing two basic extraction techniques, LLE or SPE, have been included. The reviewed

techniques are those based on spectrophotometric methods, as well as analytical

separation (gas chromatography (GC), high-performance liquid chromatography

(HPLC), and capillary electrophoresis (CE)). Many reports in the literature determine

the total amount of phenolic compounds in olive oils by spectrophometric analysis

and characterize their phenolic patterns by capillary gas chromatography (CGC) and,

mainly, by reverse phase high-performance liquid chromatography (RP-HPLC);

however, CE has recently been applied to the analysis of phenolic compound of olive

oil and has opened up great expectations, especially because of the higher resolu-

tion, reduced sample volume, and analysis duration. CE might represent a good com-promise between analysis time and satisfactory characterization for some classes of

phenolic compounds of virgin olive oils.

Key Words:Polyphenols; Olive oil; Analytical techniques; Capillary electrophoresis; CE; HPLC;GC; Spectrophotometric methods;

Received:January20, 2005; revised: March8, 2005; accepted:March 9, 2005

DOI 10.1002/jssc.200500032

1 Introduction

1.1 Importance of olive oil and its differences from

other vegetable oils

Olive oil has been produced for about 6000 years, but inthe last thirty years there has been a growing interest inthe use of olive oil in cooking because of a greater knowl-edge of Mediterranean food and an awareness of thehealthy virtues of a Mediterranean diet, and particularlyolive oil [1, 2].

Among the different vegetable oils, virgin olive oil (VOO)is unique because it is obtained from the olive fruit (Oleaeuropaea L.) solely by mechanical means, without furthertreatment other than washing, filtration, decantation, orcentrifugation [3].

Olive oil can be consumed in the natural unrefined state oras a refined product. The refined product is made eitherfrom virgin olive oil and called refined virgin oil (RVO) or

from solvent-extracted oil [4] and called refined husk oil(RHO).

Its chemical composition consists of major and minorcomponents. The major components, that include glycer-ols, represent more than 98% of the total weight. Abun-dance of oleic acid, a monounsaturated fatty acid, is thefeature that sets olive oil apart from other vegetable oils.In particular, oleic acid (18 :1n-9) ranges from 56 to 84%of total fatty acids [5], while linoleic acid (18:2 n-6), themajor essential fatty acid and the most abundant polyun-saturate in our diet, is present in concentrations between3 and 21% [6, 7].

Minor components, amounting to about 2% of the total oilweight, include more than 230 chemical compounds, e.g.,aliphatic and triterpenic alcohols, sterols, hydrocarbons,volatile compounds, and antioxidants [8]. The main anti-oxidants of VOO are carotenes and phenolic compounds,including lipophilic and hydrophilic phenols [9]. While thelipophilic phenols, among which are tocopherols, can befound in other vegetables oils, some hydrophilic phenolsof VOO are not generally present in other oils and fats [9,10].

Correspondence: A. Segura Carretero or A. Fernndez Gutir-rez, Research Group FQM-297, Department of Analytical Chem-istry, Faculty of Sciences, University of Granada, C/Fuentenuevas/n, E-18071 Granada, Spain. Phone: +34 958 243296/7Fax: +34 958 249510.E-mail: [email protected] or [email protected]


2/22

838 Carrasco-Pancorbo, Cerretani, Bendini, Segura-Carretero, Gallina-Toschi, Fernndez-Gutirrez

1.2 Phenolic compounds in VOO

Polyphenols is a broad term used in the natural productsliterature to define substances that possess a benzenering bearing one or more hydroxy groups, including func-tional derivatives [11].

According to Harborne et al. [11] phenolic compounds are

grouped into the following categories: 1. phenols, phenolicacids, phenylacetic acids; 2. cinnamic acids, coumarins,isocoumarins and chromones; 3. lignans; 4. ten group offlavonoids; 5. lignins; 6. tannins; 7. benzophenones,xanthones, and stilbenes; 8. quinones; 9. betacyanins.Most phenolic compounds are found in nature in a conju-gated form, mainly with a sugar molecule.

In the case of virgin olive oils, polyphenols mostly refersto hydrolysis products of oleuropein and ligstroside, agly-cones, and related compounds.

The phenolic fraction of VOO consists of an heteroge-neous mixture of compounds, each of which varies in

chemical properties and impacts on the quality ofVOO [12]. The occurrence of hydrophilic phenols in VOOwas observed more than 40 years ago by Cantarelli andMontedoro [13, 14]. They established a set of researchpriorities related to polyphenols which remain practicallyunchanged to this day:

Development of an analytical procedure to quantifyphenolic compounds in oils.

Estimation of the levels of phenolic compounds invegetables oils.

Possible relationship between these compounds andthe characteristics of the olive fruit (variety, degree of

ripeness). Effect of extraction technology and refining process

on the level of polyphenols.

Importance of phenolic compounds as natural antioxi-dants.

Possible role of polyphenols in justifying why olive oilswith high peroxide values have considerable stability.

It has not been easy to satisfy these points and manyresearchers are still working on them. However, veryinteresting systematic studies of the individual classes ofhydrophilic phenols in VOO have been developedrecently, and it is possible to say that the composition of

VOO is today largely elucidated.VOO contains different classes of phenolic compoundssuch as phenolic acids, phenolic alcohols, flavonoids,hydroxy-isocromans, secoiridoids, and lignans asreported in Table 1.

Phenolic acidswith the basic chemical structure of C6-C1 (benzoic acids) and C6-C3 (cinnamic acids), such ascaffeic, vanillic, syringic, p-coumaric, o-coumaric, proto-catechuic, sinapic, and p-hydroxybenzoic acid, were the

first group of phenols observed in VOO [15, 16]. Severalauthors confirmed the occurrence of phenolic acids asminor components in VOO [17 23].

The prevalent phenols of VOO, however, are the secoiri-doids, that are characterized by the presence of eitherelenolic acid or elenolic acid derivatives in their molecular

structure [24]. These compounds, e.g., oleuropein,demethyloleuropein, and ligstroside, are derivatives of thesecoiridoid glucosides of olive fruits. Breakdown productsof two major phenolic constituents of the olive fruit, oleuro-pein and ligstroside, form the majority of the phenolic frac-tion.

The most abundant secoiridoids of VOO are the dialdehy-dic form of elenolic acid linked to hydroxytyrosol = (3,4-dihydroxyphenyl)-ethanol or tyrosol = (p-hydroxyphenyl)-ethanol (3,4-DHPEA-EDA or p-HPEA-EDA) and an iso-mer of the oleuropein aglycone (3,4-DHPEA-EA). For thefirst time, these compounds were found by Montedoro etal. [20,25] who also assigned their chemical struc-

ture [26]. Later these structures were confirmed by otherauthors [27]. Recently, oleuropein and ligstroside agly-cone were also detected as minor phenolic components inVOO [28, 29].

Hydroxytyrosol and tyrosol are the main phenolic alco-hols in VOO. It is also possible to find in VOO hydroxytyr-osol acetate [30], tyrosol acetate [31], and a glucosidicform of hydroxytyrosol [32].

Several authors have reported that flavonoids such asluteolin and apigenin were also phenolic components ofVOO [33, 34]. (+)-Taxifolin, a flavanonol, has recentlybeen found in Spanish virgin olive oil [23].

The last group of phenols found in VOO are thelignans;Owen et al. [28,35] and Brenes et al. [36] have recentlyisolated and characterized (+)-1-acetoxypinoresinol, (+)-pinoresinol, and (+)-1-hydroxypinoresinol as the most fre-quent lignans in VOO. Lignans are also found as prevalentphenolic compounds in VOO.

A new class of phenolic compounds, hydroxy-isochro-mans, was found in different samples of extra-VOO. Inparticular, the presence of 1-phenyl-6,7-dihydroxy-iso-chroman and 1-(39-methoxy-49-hydroxy)phenyl-6,7-dihy-droxy-isochroman has been demonstrated [37].

1.3 The family of polyphenols in the minoritycompounds and their antioxidant, health, and

sensory properties

The antioxidant potential of phenolic compounds in oliveoil has also been a subject of considerable interest, bothbecause of its chemoprotective effect in humanbeings [38 43] and because it is a major factor in the highstability (shelf-life) of olive oils [4246]. The antioxidantactivity of VOO components has been related to the pro-



3/22

Analytical determination of polyphenols in olive oils 839

tection against important chronic and degenerative dis-eases such as coronary hearth diseases (CHD), ageingneuro-degenerative diseases, and tumours of differentlocalizations [47 49]. Among these protective effects, itpossible to highlight, for example, the protection of lowdensity lipoprotein (LDL) oxidation [50]; the reduced oxi-dative damage of the human erythrocytes by 3,4-DHPEA [51] and the reduction of free radical production inthe faecal matrix [52]. Moreover, in several studies, it wasaffirmed that the phenolic substances isolated and puri-fied from olive oil were much more potent antioxidantsthan the classical in-vivo and in-vitro free radical scaven-gers, vitamin E and dimethyl sulfoxide [28, 52 54].

Polyphenols also contribute to the organoleptic propertiesof VOOs and are commonly described as bitter and astrin-gent [44, 55 57] and responsible for organoleptic charac-teristics in general [58]. Less commonly, polyphenols areassociated with pungency, that is, peppery, burning, orhot sensations [8, 44,59]. However, the relationshipsbetween individual hydrophilic phenols of VOO and itssensory characteristics are not totally defined. Forinstance, several authors associated the off-flavour note

of atrojado with the presence of phenolic acids inVOO [60], but other studies did not show any relationbetween bitter sensory note and phenolic acid content in aVOO [61]. The relations between the secoiridoid deriv-atives and the bitterness of VOO have also been studied;first, interest was focused on two derivatives of oleuropeinand demethyloleuropein, such as 3,4-DHPEA-EDA andp-HPEA-EA [62, 63]. In this case, Garca et al. [63] stud-ied the reduction of oil bitterness by heating of olive fruits,and a good correlation between oil bitterness and contentof hydroxytyrosol secoiridoid derivatives was found. Inlater studies it was observed that a relation exists betweenthe bitter and pungent sensory properties and ligstroside

derivative content [64] the content of the aldehydic form ofoleuropein aglycone [65].

1.4 Importance and difficulties of quantification of

phenolic compounds in virgin olive oils

The qualitative and quantitative composition of VOOhydrophilic phenols is strongly affected by the agronomicand technological conditions of its production.


Table 1. Phenolic compounds in virgin olive oil and their chemical general structure.


4/22


Several agronomic parameters can modify the phenolicconcentration of VOO. The most widely studied aspectsinclude cultivar, fruit ripening, pedoclimatic conditions,and some agronomic techniques such as irrigation [6668]. As reported by different authors the phenolic compo-sition of fruits is affected qualitatively by the cultivar.

Since the occurrence of hydrophilic phenols in VOO isstrictly related to the activities of various endogenousenzymes of olive fruits, their concentration in the oil isstrongly affected by the extraction conditions. Crushingand malaxation are the most important critical technologi-cal points of the mechanical oil extraction process [6971].

For all these reasons, the identification and quantificationof the individual components of VOOare of great interest.

Many analytical procedures directed towards the determi-nation of the complete phenolic profile have been pro-posed (spectrophotometric methods; biosensors; paper

chromatography, TLC, GC with different detectors, andHPLC coupling with several detection systems; NMR andIR techniques for the characterization and identification ofthese compounds; capillary electrophoresis (CE)); how-ever, the variety of extraction techniques, chromato-graphic conditions, and methods of quantification havecontributed to differences in reported levels of virgin oliveoil phenolics.

Direct comparison between the concentrations of olive oilphenols reported in the literature is difficult, as thereported concentrations often differ greatly (sometimeseven in orders of magnitude). Several authors haveexplained this by the fact that there are numerous factorswhich affect phenolic compounds of VOO, such as vari-ous genetic characteristics of the olive cultivar [44] ortechnological modifications during processing theolives [72]. These reasons may partly, but not completely,explain these discrepancies. Pirisi and co-workers raisedthe question whether the discrepancies may be caused bythe various analytical methods used and/or the expres-sion of the results in various formats [73].

In fact, individual phenolic compounds give differentresponses during UV detection after their HPLC separa-tion [31]. Another cause of confusion is the use of differentstandard equivalent units in the case of the Folin-Ciocal-

teu colorimetric assay for total phenolics, depending onthe chosen calibration curve (e.g., caffeic acid, gallic acid,syringic acid, tyrosol, oleuropein equivalents).

It would be really interesting, as Tsimidou proposed [44],to perform a possible collaborative study using the sameanalytical method to ensure that the differences in magni-tude of the phenol content depend mainly on the variety.Two years later, Pirisi et al. [73] stated that before such astudy can be undertaken, it is necessary to study, in detail,

the influence of differences in milling conditions on thepolyphenol content of oils.

Despite a general recognition of the problems associatedwith the analysis and quantification of phenolic com-pounds in olive oil, only recently has a paper been pub-lished with the objective of highlighting the differences

between the various units used for expressing levels ofolive phenolics [74].

In general, an analytical procedure for the determinationof individual phenolic compounds in VOO involves threebasic steps: extraction from the oil sample, analytical sep-aration, and quantification. These steps will be explainedin later sections of this article.

2 Sample preparation

2.1 Hydrolysis

In a number of instances, an hydrolysis step has been

included to minimize interferences in the subsequentchromatography and to simplify chromatographic data,particularly in instances where appropriate standards arecommercially unavailable.

Acid hydrolysis has been the traditional approach tomeasurement of aglycones and phenolic acids from flavo-noid glycosides and phenolic acid esters, respectively.

However, two forces have driven the use of alkalinehydrolysis. First, commercial processing of many plant-derived foods now involves alkali-treatment and the stabil-ity of plant phenols under these conditions becomes ofinterest. For instance, the major characteristic phenols ofolive are secoiridoids and their reactivity in alkali has beenexamined [75].

Several examples showing the use of alkaline or acidhydrolysis of the polar fraction of olive oil are mentioned inthis review [18, 20, 97].

2.2 Extraction of phenolic compounds from virgin

olive oils

Isolation of phenolic compounds from the sample matrix isgenerally a prerequisite for any comprehensive analysisscheme although enhanced selectivity in the subsequentquantification step may reduce the need for samplemanipulation. The ultimate goal is the preparation of a

sample extract uniformly enriched in all compounds ofinterest and free from interfering matrix components [76].

A great number of procedures for the isolation of the polarphenolic fraction of VOO utilizing two basic extractiontechniques, LLE or SPE, have been published in the litera-ture. Various isolation systems have been proposed bydifferent authors, depending, among other aspects, onthe aim of the particular study. The systems do not varyonly in solvents and/or solid-phase cartridges used, but



5/22


also in the amount of the sample taken for analysis,volumes of solvents, and further details [74].

In addition, the phenolic fraction of VOO consists of anheterogeneous mixture of compounds, which are, in mostof cases, not commercially available. This represent aproblem in calculating the recovery of an isolation tech-

nique. Therefore, other similar phenolic compounds havebeen frequently used in the past. In recent studies, toovercome this obstacle, a refined phenolics-free olive oilis spiked with an exactly specified dose of a phenolicextract that is prepared by the extraction of VOO.

2.2.1 Liquid-liquid extraction

Traditionally, the phenolic fraction of olive oil has been iso-lated by extraction of an oil solution in a lipophilic solventwith several portions of methanol [28, 35] or methanol/water (with different levels of water that range between 0and 40% [16, 1821, 77]), followed by solvent evapora-

tion of the aqueous extract and a cleanup of the residueby solvent partition [20, 23, 30, 7883]. The most com-monly used solvent has been hexane, but petroleum etherand chloroform have also been proposed; however, theaddition of hexane or other organic solvents to the oilbefore extraction did not yield significant differences in thephenol recovery efficiency [20].

Tensioactive substances (e.g., Tween 20 (2% v/w)) haverepeatedly been used to liberate the phenolic compoundsof the lipoprotein membranes [20,84].

Extraction with tetrahydrofuran/water followed by centrifu-gation has also been performed [85], as well as extraction

with N,N-dimethylformamide [86].Montedoro et al. [20] examined the extractive methods forsimple and hydrolyzable phenolic compounds in VOO,studying different schemes where olive oil was dissolvedin various solvents, and the volume and the percentage ofthese solvents were changed. The best results wereobtained using methanol/water (80:20,v/v) in agreementwith data reported in the literature [84, 87] and optimumextraction was achieved on extracting 100 g of olive oilwith two 20-mL volumes of solvent. However, severalyears later Angerosa et al. [88] reported results that werein contrast with these; the incomplete recovery of somecomponents and the considerable emulsion formation

between the oil and the methanol/water layer promptedthem to choose chose neat methanol as the extraction sol-vent.

Cortesi et al. [85] examined the extraction of the polarfraction of olive oil with tetrahydrofuran/water (80:20, v/v)followed by centrifugation, and they concluded that therecoveries were 5 times higher with this method in termsof hydroxytyrosol and 2 times for tyrosol than with metha-nol/water (60: 40, v/v).

The used of N,N-dimethylformamide seemed to showinteresting results in terms of recovery efficiency and sam-ple manipulation [86].

After the liquid-liquid extraction process in order to isolatethe desired analytes from unsaturated, interfering spe-cies, residual oil is removed by overnight storage at sub-

ambient temperature [88], by centrifugation or by furtherextraction with hexane, although Sephadex column [20,25] and Policlar AT: Celite 560 (1:2) [17] chromatographyhave also been used to effect further clean-up.

2.2.2 Solid-phase extraction

The versatility of SPE has been exploited for the recoveryof phenolic compounds from olive oil and various systemsemploying SPE, either as isolation or clean-up step, havebeen reported in the literature.

Some of the suitable sorbents are alkylsilicas, such asC8[73, 89] or C18[22, 90](but incomplete extraction of the

phenolic fraction [91] and partial oil separation [73] havebeen reported). Despite the common assumption that C18phase is less suitable for the isolation of polar componentsfrom a non-polar matrix than normal-phase SPE, C18-car-tridges have often been tested for isolation of phenolsfrom VOO [9294]. Anionic exchange cartridges havebeen also used to isolate the phenolic fraction from vari-ous seed oils, but recoveries were low (5362%) forsome components [95]. Promising results were obtainedby Mateos et al. [31], who worked with amino-phase car-tridges and diol-bond phase SPE cartridges and found forthe latter high recovery (>90%) of all major olive phenoliccompounds.

In doped refined olive oil samples the recoveries werestudied using a C18-cartridge with total supression ofresidual silanol groups (C18EC, end capped) [96]. In thisstudy the discrimination between C18 andC18EC was not sovast as expected since the two stationary phases differonly in the presence of free silanol groups; however, theirpresence (C18) seems to improve the release mechanism,increasing the recovery.

Anyway, in all the cases, one of two experimentalapproaches was used. In one procedure, a solution of theoil in hexane was applied [22] to a pre-conditioned (typi-cally reversed-phase) cartridge which was washed withhexane-ethoxyethane or hexane-cyclohexane [89] mix-tures to remove the non-polar lipid fraction. Phenols werethen eluted with acetonitrile or methanol. Alternatively, thepolar fraction of olive oil has been partitioned into aqueousmethanol from a hexane solution [97] and fractionatedinto two parts (A and B) by SPE. Analysis of the two frac-tions show that Part A (eluted from Sep Pak C18 withmethanol/water (50 :50v/v)) contains only simple phenolsand phenolic acids, while Part B (eluted with mixtures ofmethanol/chloroform) has a complex nature.



6/22


It is possible to find some papers where the authors use acombination of LLE and SPE in the same sample prepara-tion protocol. For instance, Buiarelli and co-workers [98]extracted phenolic acids of olive oil with borate buffer atpH 9.2. Then the extraction buffer solution was loaded ona phenyl cartridge, and the eluent of this step was acidified

to pH 2.6 and loaded on a C18cartridge. Plant phenols areionizable with typical pKavalues ranging from 8 to 12 andoil/water partition coefficients ranging from 6610 4 to 1.5.Thus, they exhibit considerable diversity in terms of acidityas well as polarity, ranging from hydrophobic to hydrophil-ic in character. The range of physicochemical behaviourmust be considered when determining sample handlingstrategies such as, for example, in pH control to ensurefavourable partitioning behaviour during extraction [99].Another example of a similar process has beenreported [100]. In a recently published paper, two differentmethods were necessary for the complete quantificationof phenolic compounds, a methanolic extraction and anextraction which included a SPE cleaning step [101].

Summarizing all the points raised so far in this section, wecan say that a reasonable performance of C18-phasecould generally be demonstrated for simple phenols; how-ever, poor recovery has been found for secoiridoid deriv-atives, especially dialdehydic forms [22, 86, 92, 93]. In arecently published study [102], a group of 15 phenoliccompounds (one of which occurs naturally in VOO) wereextracted in order to study the recoveries of each analyte,and this work showed that LLE gave the best results interms of recovery of the phenolic standard mixture addedto the refined peanut oil. In the same study, diol-SPEshowed higher recoveries of total phenols, o-diphenols,

tyrosol, hydroxytyrosol, and secoiridoids than the otherextraction procedures (C8-SPE; C8 mod.-SPE; and C18).This is in accordance with another published work [74],where the experimental work demonstrates that the appli-cation of LLE led to significantly recovery of total phenoliccompounds (93%) than SPE-diol (68%) and SPE-C18(38%). These total averages were calculated by spiking arefined phenolics-free olive oil with an exactly specifieddose of a phenolic extract that was prepared by theextraction of VOO.

2.2.4 Supercritical fluid extraction

Supercritical fluid extraction (SFE) was developed in the1960s and, in recent years, has acquired some relevancefor the extraction of polyphenols from plant sources. Themain advantage of SFE is that it combines the character-istics of gases and liquids for extraction. The low viscosityof the supercritical fluids confers a high capacity for diffu-sion and improves access to phenolic compounds boundto the cell wall [103]. Moreover, their relatively high den-sity confers a high solvation power, which greatly facili-tates the extraction process. Furthermore, it minimizes

any possible degradation processes [104], such as oxida-tions or isomerizations, that may occur with other moreconventional extraction techniques, because it reducesextraction time and because the process can be carriedout in the absence of light and air.

The extraction behaviour of phenolic compounds covering

a range of polarities has been modelled using supercriticalcarbon dioxide and an inert support as a samplematrix [105]. Extraction and collection variables were opti-mized and revealed that the use of methanol as modifierwas mandatory. Dynamic SFE produces clean extractswith higher recoveries of total phenols from dried oliveleaf [106] than sonication in liquid solvents such asn-hex-ane, ethoxyethane, and ethyl acetate. However, theextraction yield obtained was only 45% of that obtainedwith liquid methanol.

The use of this extraction system is still rare for the extrac-tion of phenolic compounds of olive oil; perhaps futurestudies will make this extraction system more common.

3 Spectrophotometric determination:advantages and disadvantages

One method that is widely used for the quantitative deter-mination of total phenols in VOO is colorimetric assay,based on the reaction of Folin-Ciocalteu reagent with thefunctional hydroxy groups of phenolic compounds [107,108]. The method consists of calibration with a pure phe-nolic compound, extraction of phenols from the sample,and the measurement of absorbance after the color reac-tion. The popularity of this assay can mainly be attributedto its simplicity and speed of analysis [109].

The major disadvantage of the colorimetric assay is itslow specificity, as the color reaction can occur with anyoxidizable phenolic hydroxy group. Recently an interest-ing approach to the content of total extractable phenoliccompounds in different food samples involving compari-son of chromatographic and spectrophotometric methodshas been reported, accounting for the possible influenceof other substances as interfering compounds [110].

Furthermore, the method does not distinguish betweenindividual compounds differing in molar mass (rangingfrom 138 to 416 g/mol in the case of the major olive oilphenols) and structure (i.e. the number of active hydroxy

groups) [74].Thus, samples with comparable total phenolic content,but widely varying phenolic composition, will give a differ-ent response in the colorimetric method.

However, Singleton et al. [111] have shown that the molarabsorptivity per reactive hydroxy group is comparable incompounds with otherwise similar structure. The molarabsorbance of an olive phenol therefore primarilydepends on the number of hydroxy groups, with mono-



7/22


8/22


4 Chromatographic determination of thephenolic profile of virgin olive oils

The need for profiling and identifying individual phenoliccompounds necessitates the replacement of traditionalmethods by high-performance chromatographic anal-yses. Separation is commonly achieved by HPLC,

although GC is used in some instances. The most com-mon mode of separation exploits reversed-phase systemstypically with C18 columns and various mobile phases.Detection is routinely achieved by ultraviolet absorptionoften involving a photodiode array detector although theversatility of the latter often appears to have beenneglected.

Coupled techniques, particularly various mass spectralmethods, are being used increasingly for routine workalthough analyte collection using preparative-scale HPLCand off-line identification are often still needed for non-routine samples.

4.1 Thin-layer chromatography

Paper and thin-layer chromatography (TLC) methodshave been devised and exploited for the preliminary sep-aration/clean-up of phenolic extracts of VOO.

In the past, paper chromatography had limited applicationto the separation of phenolic compounds of VOO [120].However, the separation of approximately 20 phenoliccompounds or several group of phenols from olive oilusing TLC on silica, cellulose, and polyamide has beendescribed. Elution of the compounds of the polar extractof olive oil has been done by dimensional-TLC usingdifferent systems:

n-Butanol/acetic acid/water (4: 1 : 5) (System A) and water(System B); toluene/ethyl formate/formic acid (5: 4 : 1)(System A) and acetic acid (2%) (System B); benzene/methanol/water (80: 1: 1) (System A) and benzene/methanol/water (45: 8 : 4) (System B) [15, 84].

Ragazzi et al. [121] used TLC with cellulose and poly-amide as stationary phases. In the first case, the mobilephase was as follows: chloroform/acetic acid/water(8:2: 1) or n-butanol/acetic acid/water (6:2:1), while inthe second case, they used methanol/acetone/water(3: 1: 1) or methanol/acetone/water (6: 1: 1).

4.2 GC and GC-MS analysis of phenoliccompounds from olive oil

The qualitative and quantitative determination of phenoliccompounds from olive oil can be accomplished by capil-lary gas chromatography (GC) of these compounds ortheir derivatives. GC in general assumes that the com-pounds injected are volatile at the temperature of analysisand that they do not decompose at either the temperatureof injection or that of analysis. In standardized analytical

methods, flame ionization detection (FID) is most widelyused. Mass spectrometry (MS) allows acquisition ofmolecular mass data and structural information, and iden-tification of compounds.

The first paper about the separation of phenolic com-pounds of olive oil by GC was publishedby Janer del Valle

et al. [122] 25 years ago. At the same time, this techniquewas also used by Solinas et al. [77,123], who used GC foridentification of mixtures between virgin olive oils andrefined oils. In 1987, Forcadell et al. [124] developed aprotocol for the preparation of trimethylsilyl (TMS) deriv-atives. In the same year, Solinas [19] published a paperon development of a GC method for the qualitative/quanti-tative evaluation of phenolic compounds in VOOs fromdifferent cultivars at diverse degrees of ripeness. In thiswork, some phenolic compounds were found in all casesalthough their amounts varied in each variety. Ratiosbetween some of these phenolic compounds seemed tobe constant during olive maturation, so that their levels

could be used as varietal markers. The methodologyinvolved liquid-liquid extraction of the phenolic com-pounds, clean-up of the methanolic extract followed by anazeotropic distillation to remove the solvents, low-pres-sure column chromatography to clean-up the extracts,and finally capillary GC analysis of the trimethylsilyl deriv-atives. Although the method permitted characterization ofthe simplest compounds, other linked phenols occurringin large quantities were not identified.

Improvement in the identification of compounds wasobtained with sophisticated analytical techniques such asGC-MS [79, 125] and GC-MS/MS [126]. For instance,

Angerosa et al. [88] used a modified extraction procedurefollowed by capillary GC-MS to identify the simple and thelinked phenols present in VOO.

Compared with direct inlet mass spectra, the GC-MSdata, in general, exhibit the same typical fragmentationpatterns but with slight differences in intensities. Ininstances where the classical mass spectrometric gasphase ionization techniques such as EI and CI are unsui-table (e.g., with polar, non-volatile, and thermolabile phe-nols), chemical derivatization usually involving silylationmay overcome these limitations but can introduce furtherdifficulties by increasing the molecular mass of the analyte

possibly beyond the range of the mass analyzer [104].Derivatization also often produces mixtures of partiallyderivatized compounds [27] from a single analyte.

Nevertheless, in the last years, the qualitative/quantitativeanalysis of the phenolic profile of VOO by GC has beenused both in analytical [96] and in applied [79,126128]work; however, the other chromatographic methods aremore often employed because they avoid the use of deri-vatizing reagents. Another problem of this technique, as



9/22


mentioned before, is that the use of high temperaturecould damage the analytes. Table 2 shows some GCtemperature ranges used, type of derivatization, charac-teristics of the column employed, and several observa-tions about several published works, as well as the time ofanalysis in each case. We have included all the GC refer-

ences mentioned during this review and two new refer-ences [129, 130].Figure 1depicts an example of a chro-matogram of phenolic extracts of a VOO, showing the pre-sence of interferences represented by the TMS deriv-atives of some fatty acids.

4.3 HPLC and LC-MS analyses

The limited volatility of many phenols has restricted theuse of GC for their separation, so HPLC currently repre-

sents the most popular and reliable technique for analysisof phenols.

HPLC is normally used for separating non-volatile, high-molecular-mass constituents, in either adsorption or parti-tion mode. Adsorption chromatography, namely normalphase chromatography, is widely used to separateclasses of constituents according to the nature and num-ber of polar functional groups [131]. In normal phaseHPLC the absorbent is silica gel and the eluent is a non-polar solvent. Initial experiments were done using thismode [132], but in the same period the best results, interms of reproducibility of retention time and separation ofthe most polar compounds, were obtained using reverse-phase HPLC [60, 123, 133, 134].


Table 2. GC conditions for determination of phenolic compounds in olive oil.

Temperaturerange

Column c haracteristics Analysis t ime Derivatization Detector s ystem Observations References

Progam 1:1602508CProgram 2:1302508C

Capillary column SE-30,3%on Chromosorb W HP2m,4.0mmID80 100 mesh

120 min Trimethylsilylation(Use of 3 different compounds:TRI-SIL (HMDS/TMCSin Pyri-dine(2:1:10);TRI-SIL/BSAP(BSAand Pyridine);TRI-SILZ(TMSI in dry Pyridine))

FID TY, H YTY a nd p henolic a cids Janer d el V alle e tal. [122]

1402608C OV 1 7, O V 1 01 a nd S E30(1:1:1)3%onChro-mosorbW HP; 100 200mesh

120 min Silylation FID Phenolic acid and simple phe-nols. TY,HYTYand 4-hydroxy-phenylaceticacid representedabout 50%of total phenolcon-tent.HPLCanalysis too

Solinas et al.[77, 123]

602708C Capil larycolumn25m,0.3mmID

45 min Silylation FID; MS Use of TLC for separating polarfraction of olive oil

Cortesi et al.[18]

702758C Capillarycolumn SE5215m, 0.15lm ID

107.5 min Trimethylsilylation FID Study of relation between oxida-tion products and negative orga-noleptic note

Solinas et al.[19]

702758C Capillarycolumn SE5425m, 0.32lmID0.10 lmfilm thickness

108 m in Trimethyl si lylat ion FID; M S; N MR The p resence o f a l igstrosideaglycon containing no carbo-methoxy groupand oleuropeinaglycone derivativeswas evi-denced

Angerosaet al.[88]

403208C DB5 MSCapillarycolumn

30m, 0.25lmID0.25 lmfilm thickness

L60min Trimethylsilylation

(bis(trimethylsilyl)trifluoroacet-amide(BSTFA))

MS(CI);NMR Phenolicandsecoiridoidsagly-

cones

Angerosaet al.

[27]

70280 8C Supelcosilica capillarycolumnSE-54;30 m,0.25 lmID0.25 lmfilmthickness

50min Trimethylsi ly lation(Derivativeswere produced withBSTFA)

MS Study of the evolution of pheno-liccompounds in VOOduringstorage. HPLCanalysis too.

Cinquantaet al.[66]

70270 8C SPB-5 fused-silica capil-lary column; 30 m,0.32 lmID0.10 lmfilmthickness

110 min Trimethylsilylation(bis(trimethylsilyl)trifluoroacet-amide)

FID;MS Comparat ivestudyofSPEusingC18and C18EC. TY,HYTY,phenolic acidsand oleuropeinaglycone and ligstrosideagly-cone.

Liberatoreet al.[96]

703008C HP1 capillary column(Hewlett-Packard) of30 m60.32mmID0.10 lm film thickness

L95min Trimethylsilylation(trimethylsilyl etherswere ob-tainedwith a silylation mixturemade up of pyridine,hexa-methyldisilazane and trimethyl-chlorosilane(2:1:1))

FID;MS (ES) HYTY, TY, oleuropein, oleuro-pein aglycones and oleoside-11-methyl esterin olivefruits

Marsilio et al.[129]

60275 8C J&WDB-5MScolumn,30 m60.25mmID0.25 lm film thickness

L30min Der ivat izat ionwithBSTFA: trimethylchlorosilane(TMCS)(99: 1)

MS; MS/MS Phenoliccompounds inSicilianoliveoils. LLE (methanol/water(80:20 v/v)).TY, HYTY,decarbomethoxyligstrosideand oleuropein agly-conesin thedialdehydic formswere themostabundant com-pounds

Saittaet al.[130]


10/22


Reverse-phase HPLC, which is based on partition chro-matography, is used to separate individual componentsthat belong to one constituent class [131]. In this case, thestationary phase usually consists of a non-polar octade-cylsilane (C18) bonded phase, while the mobile phase is apolarsolvent.

Columns are commonly 100 to 300 mm in length with10lm or 5 lm packings, shorter columns being givenpreference.

In some instances, isocratic elution has provided ade-quate resolution due to the selectivity effects of one ormore components (e.g., acetonitrile) of the mobile phase,although gradient elution has usually been mandatory inrecognition of the complexity of the phenolic profile ofmost samples. Numerous mobile phases have beenemployed, but binary systems comprising an aqueouscomponent and a less polar organic solvent such as acet-onitrile or methanol remain common. Acid (acetic, formic,or phosphoric acid) is usually added to both componentsto maintain a constant acid concentration during gradientruns. The decrease of pH helps to avoid the dissociation

of phenolic compounds, improving also the asymmetry ofthe peak and reducing the peak-tailing [135].

For example, a method available for analyzing the pheno-lic fraction is reversed-phase HPLC using isocratic elutionwith an aqueous solution of sulfuric acid-acetonitrile [89]or with methanol-aqueous acetic acid [136]. The tech-nique, as stated before, has also been applied using gra-dient elution with methanol-aqueous acetic acid [80] oracetonitrile-aqueous acetic acid [137].

Routine detection in HPLC is typically based on measure-ment of UV-Vis absorption at 225, 240, or 280 nm.Because some phenolic compounds show severalabsorption maxima, the use of simultaneous multiple UV(photodiode array) is recommended for identification pur-poses [20, 67, 89]. Indeed, there are significant differ-ences in absorption maxima and molar absorptivities [80]of even the major phenols identified in a single olive fruit.This creates problems in quantification as discussed byTsimidou et al. [80] who classified the various phenolsinto four groups and used a single calibration standard forthe members of each group. The most commonly usedwavelength for routine detection has been 280 nm whichrepresents a suitable compromise, although detection atother wavelengths and dual wavelength [20] detectionhave been applied. The advantages of low wavelengthdetection at 225 nm have been demonstrated [89] butproblems associated with high background absorption oftypical mobile phases in RPLC have limited its use.

Identification of the eluted phenols in HPLC (and in GC) isusually based on correspondence of the retention datawith an appropriate standard. The limited availability ofsuitable standards for quantification is a problem that canbe overcome, in part, by synthesis of the relevant com-pounds. Alternatively, the relevant compounds isolated bypreparative scale chromatography can serve as referencestandards [31, 35, 59].

In many instances, quantification is carried out by refer-ence to one or more appropriately selected referencecompounds.


Figure 1.Example of a GC chromatogram of a phenolic extract of virgin olive oil. Key: I.S. resorcinol (internal standard); 1, tyro-sol; 2, hydroxytyrosol and phenolic acids zone; 3, C16:0fatty acid TMS derivative; 4, C1 8 : 0, C1 8 : 1, and C1 8 : 2fatty acids TMS deriv-

atives;5, ligstroside aglycone zone; 6 , monoglyceryl TMS derivatives zone; 7 , oleuropein aglycone zone. From reference [96]with permission of the authors.


11/22


However, all phenols possess a strong chromophore sys-tem. Their UV spectra are particularly informative, provid-ing considerable structural information that can distin-guish the type of phenol and the oxidation pattern.Furthermore, spectra of eluting peaks obtained at, forexample, the apex and the two inflection points of the

peak can be compared and used as indicators of purity.The complementary nature of fluorescence detection hasbeen demonstrated. In fact, the lignans show a goodresponse to fluorescence excitation and this techniquehas been studied by Brenes et al. [138] to analyze thesephenolic compounds in olive oils. Although fluorescencedetection offers some advantages over UV detection interms of enhanced selectivity and sensitivity, not all thephenolic compounds found in food fluoresce [139]. Theanalysis of phenolic compounds in olive fruit using fluores-cence detection has also been reported [140], but there isonly one paper reporting the analysis of some cinnamicacids in olive oils.

Coulometric electrode array detection coupled to HPLCwas also used for qualitative and quantitative determina-tion of phenolic compounds in VOO [86, 138]. The advan-tage of this method, besides easy sample preparation, isthe possibility of separating and quantifying co-eluted sub-stances with different potentials [141, 142]. The detectormeasures the different potentials at which the phenoliccompounds are oxidized [143]. In addition, an ampero-metric detector was also used to quantify phenolic com-pounds in olive oils, showing great sensitivity and stabilityalthough the contamination of the electrode with the oxi-dation products is an important disadvantage [144].

Despite the obvious successes of GC-MS, as was com-mented before, it is the hyphenation of liquid chromato-graphy with MS that has revolutionized the analysis ofnon-volatile species, as evidenced by the number ofpapers.

LC-MS interfacing has been achieved in a number ofways; however, it was with the advent of API techniquesthat LC-MS came of age. API is a soft ionization sourcefor the analysis of polar, non-volatile, thermolabile, andhigh molecular mass molecules such as plant phenols(e.g., phenolic compounds of olive oil [145] and oliveleaf [146]). Although API has revolutionized the applica-tion of LC-MS, some problems remain, the major limitation

being the strong dependency of the response on the nat-ure of the analyte plus the mobile phase [104]. Thus, gen-eration of mass spectral libraries is difficult. Moreover, it isdifficult to optimize conditions for a typical extract contain-ing a broad range of analytes although many instrumentsnow have provision for programmed operation of spectralconditions.

API mass spectra typically comprise protonated molecu-lar ions, [M+H]+ or sodium adduct ions [M+Na]+ in positive

ion mode, or deprotonated molecular ions, [M H] innegative ion mode with few fragment ions and thus have alow structure information content. On rare occasions, LC-MS can provide data sufficient for full structure analysisbut more generally it is used to determine molecular massand to establish the distribution of substituents on the phe-

nolic ring(s).HPLC with tandem mass spectrometry (LC-MS-MS) andnegative APCI was used for the analysis of phenolic acids,tyrosol and oleuropein derivatives [147].

Negative ESI was more sensitive for the majority of phe-nols. Thus, molecular masses of the separated phenoliccompounds were obtained through prominent [MH]

ions for most of the compounds.

There are several examples of papers where MS or MS/MS is used for the analysis of phenolic compounds in oliveoils and olives (see Table 3).

In instances where mass spectral data are insufficient to

establish a definitive structure, NMR spectrometry is apowerful complementary technique for structural assign-ment. NMR spectra of phenolic compounds are frequentlycomplex and identification of the isolated compounds iscomplicated in the absence of suitable reference stan-dards which requires time-consuming syntheses of therelevant materials. Although 2D NMR spectrometry canbe used for structural analysis without a reference com-pounds, the technique requires relatively large amounts ofthe compounds. Limited sensitivity and the need to isolaterelatively large quantities of sample are currently thegreatest limitations of NMR spectrometry.

In many instances, the combination of UV, MS, and 1

HNMR will provide adequate information for structural eluci-dation [28, 31, 35, 74]. In other cases, information on the13C NMR signals is necessary plus 2D correlations experi-ments such as COSY or 1H-13C correlation experimentssuch as HMBC or HSQC.

Several authors have analyzed phenolic compounds inolive oil, olive oil waste water, and olive fruits using differ-ent methods of HPLC, with different extraction systemsand coupling diverse detector systems. All this informa-tion is summarized in Table 3, where the time of analysis,the type of elution employed, the mobile phases, the sta-tionary phase, the extraction system, the detection sys-

tem used, and several pertinent observations are given.We have included all the HPLC references mentionedduring this review and several new references [148 157].Obviously, it is impossible to list all papers in this table.We have therefore chosen several of the most represen-tative papers.

For all the reasons explained before, HPLC is the tech-nique most commonly employed for analysis of the polarfraction of olive oil. However, GC and HPLC could be



12/22



Table 3. Separation of phenolic compounds of polar fraction of olive oil using HPLC methods

Timeofanalysis

Mobi lephases Stationaryphase Typeofelut ion Extractionsystem Detect ionsys-tem

Observati ons References

45 6 0 m in MeOH: H2O (50:50 v/v)+ CH3COOH0.2%

lBondapak C18,30cm63.9mmID10 lm

Isocratic LLEwithmethanol/water(60:40 v/v) (30gof olive oil)

UV(280 nm) Study ofdifferent Span-ishvirgin olive oilsob-tainedfrom olives atdifferent degree of ripe-ness

Graciani-Cost-anteet al.[60,133]

60 min A: H2O + C H3COOH2%B: MeOH

Hibar RP18 Gradient LLEwithmethanol/water(60:40 v/v)

UV Analysis of virgin, re-fined andsolvent ex-tracted oliveoils

Cortesi et al.[134]

60 min A: H2O + C H3COOH0.2%B: MeCN

lBondapak C18 Gradient LLE w ith methanol UV Tentative i dentificationof mixtures betweenvir-ginoliveoilsand refinedoils

Solinas etal.[123]

70 min A: H2O + C H3COOH2%B: MeCN:MeOH(50:50v/v)

Hibar RP18,25cm64.6mmID7lm orSpheri-sorbODSC18,25cm64.6mmID5 lm

Gradient LLE UV Use of alkaline or acidhydrolysis ofthe polarfraction of oliveoil

Cortesi et al.[18]

70 min A: H2OB:MeOH

UltrasphereODS C18,25cm62.0mmID5 lm

Gradient SPE (C18) UV (225 nm) Correlation of bittertaste ofVOO andinstru-mentalHPLC analysis

Gutirrezet al.[90]

45 min A: H2O + C H3COOH2%(pH3.1)B: MeOH

Erbasil C18,15cm64.6mmID

Gradient LLEwithmethanol/water in differentpro-portions

UV Use of TLC.Elenolic acid and4 un-knowncompounds(239nm).Phenolic acidsand

secoiridoids.

Montedoro etal.[20]

HPLC method of Montedoro et al. [20] Gradient LLE with methanol/ water(80:20 v/v)eva-poration to syrup con-centration;addition ofacetonitrile;washingwith hexane.

UV Characterization ofoleuropein aglyconeand 3 hydrolyzable phe-nols.Analysisof pheno-licacidsand secoiri-doids.Useof hydrolysisofphenolic extracts, TLCandHPLCfor doingstandard compounds

Montedoro etal.[25]

HPLC method of Montedoro et al. [20] Gradient LLE with methanol/ water

UV;NMR andIR

Spectroscopic charac-terization of secoiri-doidsderivatives

Montedoro etal.[26]


SpherisorbODS 2,25cm64.6mmID5 lm

Gradient Aqueousethanolex-traction LLE

UV Calculation of molaradorptivitiesof indivi-dual compounds.Cali-bration with 4 standard

curves in groups(130 min! segmentsof mobilephase)

Tsimidouet al.[80]

50 min A: H2O + C H3COOHpH3.2B: MeOH

Erbasil C18,25cm64.6mmID10lm

Gradient LLEwithmethanol(30gofoliveoil)

GC-MS; MNRanalysis

Developmentof HRGCmethodof phenoldo-sage.Characterization of sev-eral complex moleculesbymeansof GC-MS

Angerosaet al.[88]

130min A: H2O + C H3COOH2%B: MeOH

HPLC-DAD:Lichro-spher100 RP18,25cm64.0mmID5 lm

Gradient LLEasVzquezRon-cero et al.[34](50g olive oil)

Variable wave-lenghtUV de-tectors (280and285 nm);electrochemi-cal

Variable wavelengthUV detectors werefound to bemore suita-blethan DADfor quanti-tativeinformation

Tsimidouet al.[21]

40 min 10 3 M H2SO4 andCH3CN

ODS-1,ODS-2, C8, C1,CNand Phanylanalyti-cal columns (Spheri-sorb, 25 cm64.6mm

ID 3 lm)

Isocratic SPE ( C8) DAD (k =225nm)

Lowquantities of oliveoilused inthe extractionsystem(1 g).Simple and complex

phenols on different col-umns.

Pirisietal.[89]

120min A: H2O + C H3COOH3%B: MeOH

SpherisorbODS,25cm64.6mmID10 lm

Gradient LLEandSPE;acidandbasichydrolysis

Colorimetry Two partof polar frac-tion=A: simple phenolsandphenolicacids; B.complex nature

Litridou etal.[97]


SpherisorbODS 2,25cm64.6mmID

Gradient LLEwithmethanol/water(80:20 v/v)

UV;MS (ESI)in positive ionmode

Flavonoids such aslu-teolinand apigeninwere detected asphe-noliccomponentsofVOO

Rovelliniet al.[33]


13/22



Table 3. Continued.

Timeofanalysis

Mobi lephases Stationaryphase Typeofelut ion Extract ionsystem Detect ionsys-tem

Obse rvati ons Ref erences


lBondapak C18,15cm64.6mmID

Gradient Angerosaetal . [88] Photodiodearray(k =280nm)

Studyof theevolution ofphenolic compounds inVOO duringstorage.GC-MSanalysistoo

Cinquanta etal.[66]

25 min A: H2O (adjustedpH 2.5

by H3PO4)B: MeCN

Supelcosil ABZ+ Plus,

25cm64.0mmID5 lm

Gradient 100g offrozenpulp

wereextractedas Amiotet al.[157]

Photodiode ar-

ray

Demethyloleuropein

as possible varietalmarker.Elenolic acidglucosideandHYTYas indicatorsof maturationfor olives.

Estiet al. [67]

100min A: H2O (adjustedpH 3.2by H3PO4)B: MeCN

LichrosorbRP18,25cm64.6mmID5 lm

Gradient Aqueousethanolex-traction with bisulfite;hexanepatitioningandSPE.(LSE procedurewith Extrelut cartridge)

HPLC-MS;HPLC-DAD

Olivefruit.Verbascoside, antho-cyaniccompounds adnoleuropein derivatives.HPTLC

Romaniet al.[148]

60 min A: H2O + C H3COOH0.2%B: MeOH

SpherisorbODS 2,25cm64.6mmID

Gradient SameasMontedoroetal.[20]

Photodiode ar-ray; MS; NMR

Simple phenols, flavo-noids,secoiridoids

Breneset al.[30]

HPLC method of Montedoro et al. [20] Various, e.g., aqueousmethanol extractioncontaining diethyldithio-carbamate followed bySPE

HPLC-DAD Olive fruit,virginoliveoil, vegetation waters,and pomace

Servilli et al.[93]

70 min A: H2O + C H3COOH2%B:MeOH+CH3COOH2%

ColumnRP18 Peco-sphere; 8.3cm64.6mmID3 lm

Gradient LLEwithaqueousmethanol

DAD Bitterness i ndex K 225and autoxidationstabil-ity

Beltrn et al.[82]

HPLCmethod ofMontedoro etal.[20] ColumnRP18 Latex;25cm64.0mmID5 lm

Gradient LLEwithmethanol(500g ofolive oil)

UV;MS (ESI)in negative andpositive ionmode;NMR

Identificationof lignansasmajorcomponents inpolar fractionof olive oil.Preparative thin-layerchromatography (PLC).

Owenet al.[35]

HPLC methodof Montedoroet al.[20,25,26]

ColumnRP18 Latex;25cm64.0mmID5 lm

Gradient LLEwithabsolutemethanol and metha-nol/water (80:20v/v)

UV;MS (ESI)in negative andpositive ionmode;NMR

Use ofTLC,GC,GC-MSStudyof antioxidant/an-ticancer capacity (Owenet al.[27])(using thesamemethod)

Owenet al.[28]

65 min Method UV: HPLCmethodof Breneset al.[30]MethodEC: A:30mM

LiClO4 solution (pH3.1with HClO4)B: MeOH containing30mM LiClO4 solution


Gradient LLE:Montedoro etal.[20];SPE: Favatiet al.[92];New extraction method:

N,N-dimethylformamide(DMF)

UV; Electro-chemical de-tector(EC)

Treatment of extractedoilwith2NHCltochecktheeffectiveness of theextraction methods.

Breneset al.[86]

25 min H2O:CH3CN (82:18v/v)+ CH3COOH0.02%

Nucleosil ODS,25cm62.1mmor25cm61.1mmID5 lm

Isocratic LLEwithbuffer;SPEwith phenylcartridges(acidification)

UV, Spectro-fluorimetric,MS, MS/MSHPLC-APCI(negative ionmode)

Phe nolic a cids Cartoni e t a l.[100]

Method1:40 minMethod2:30 min

Method1: HPLC meth-odof Pirisiet al. [89]Method2: A: HPLCmethodof Montedoroetal.[20]


Gradient LLEandSPEsystems UV;DAD Simpleandcomplexphenols

Pirisiet al.[73]

HPLC methodand conditionsof Cortesietal.[85]

C18 column (RP)Alltech25cm64.6mmID

Gradient LLE:Montedoro etal.[20] usingbutylatedhydroxytoluene (BHT)

MS; MS/MS Analysisof oleuropeinaglycone by APCI-MS.Phenolic compoundprofile

Carusoet al.[149]

HPLC method and conditions of Brenes et al. [30] Photodiodearray

Enzymes duringma-laxation

Garcaet al.[81]

50 min A: H2O + C H3COOH1%B: MeOH/MeCN/CH3COOH (95:5:1v/v/v)

C18 column (RP)15cm64.6mmID5 lm

Gradient Aqueousmethanol ex-traction

UV;APCIandESI (negativeand positiveionmode)de-tection

OlivefruitSemiprepartive scaleHPLCtoo; antioxidantactivity studies

McDonald et al.[117]

HPLCmethod ofRomaniet al. [148] LichrosorbRP18,25cm64.6mmID5 lm

Gradient LLEwithEtOH/water(70:30 v/v), thewaterwasacidifiedwithfor-micacid(pH 2.5)

DAD;MSD HPLCanalysisofphe-nolicacids, secoiridoidsand flavonoids

Romaniet al.[127]


14/22



Table 3. Continued.

Timeofanalysis

Mobi lephases Stationaryphase Typeofelut ion Extractionsystem Detect ionsys-tem

Observati ons References


UltrasphereODS C18,25cm64.6mmID5 lm

Gradient According tothemeth-oddescribedby Capo-nio etal.[42]

UV(278 nm) Influenceof the degreeof olive ripenesson or-ganoleptic characteris-tic and shelf-life(VOOfrom Coratinaand Ogliarola salentinacultivars)

Caponioet al.[43]


ColumnRP18 Symetry10cm64.6mmID3.5lm

Gradient Olive oil residuesad-justed to pH 3 with HCland extracted with ethylacetate.

UV/Vis Oliveoi l residues.Sim-ple phenolic com-pounds

Lesage-Mee-sen etal.[150]

100min A: H2O (adjustedpH 3.2by H3PO4)B: MeCN

LichrosorbRP18, 25cm64.6mmID5 lm

Gradient Acidificationof wastewatersand LSE(Extre-lut cartridge).Elutionsteps:1) hexane;2)ethyl acetate; 3)acidmethanolExtractionof olivepulp(asRomani et al.[148])

DAD;MS Oliveoi lwastewaterandrelated olivesam-ples

Mulinacciet al.[151]

50 min A: H2O + C H3COOH3%B: MeCN:MeOH(50:50v/v)

Lichrospher100 RP18,25cm64.0mmID5 lm

Gradient ComparativestudiesofLLEand SPE usingdiol-phase cartridges;unwanted substanceswashedout withhexane

and hexane/ethylacet-ate (90:10, v/v)

UV;DADNMR(forlig-stroside agly-cone)

Phenols, flavones andlignans.Colorimetric determina-tionof o-Diphenols.GC-MS

Mateoset al.[31]

70min Elut ionsolventsused:A: H2O + C H3COOH1%B: MeOHC: AcetonitrileD: Isopropanol

Apex octadecyl104 C18;25cm60.4mmID5 lm

Gradient LLEwithmethanol andisopropanol/methanolmixture

UV-Vis Phenol iccompoundsand tocopherols(280nm)(simpleand com-plex phenols anda-to-copherol)

Tasioula-Mar-gari et al.[152]

60 min A: H2O + C H3COOH2mMB:MeOH+CH3COOH2mM

Nucleosil ODS,25cm62.1mmID5 lm

Gradient LLEwithmethanol/water(80:20 v/v), acidi-fication and passedthrough a C18 cartridge

MSandMS/MS(API/MSinnegative ionmode)

Identificationof a newclassof phenolic com-poundsin olive oils: hy-droxy-isochromans

Biancoet al.[37]

Method UV: HPLC method of Brenes et al. [30] Gradient L LE with N,N-dimethyl-formamide (DMF)

DAD;fluores-cence

Phenolic compounds inPicual variety

Garcaet al.[153]

HPLC method and conditions of Brenes et al. [30] UV; electro-chemical, fluor-escence, MS.

Useof a lignan (1-acet-oxypinoresinol) toauthenticatePicualoliveoils.UseofGCtoo.

Breneset al.[138]

65 min A: H2O + C H3COOH5%B: MeOHC: MeCN

Spherisorb S3ODS 2,25cm64.6mmID5 lm

Gradient SameasMateosetal .[31]

DAD(240,280,335nm)

Phenolic acids,secoiri-doids,lignans in Corni-cabravirginoliveoils(50min ofseparation +15minto clean the col-umn)

Gmez-Alonsoet al.[154]

50minor70min

A: H2O + Phosphoricacid0.5%B: MeOH/MeCN (50:50v/v)


Gradient SPE(diol-boundphase) UV,HPLC-MS inESI(positiveion mode)

Dialdehydicand aldehy-dicformsof oleuropeinaglycone and ligstro-side aglycone

Gutirrez-Ro-saleset al.[56]

65 min A: H2O + C H3COOH2%B: MeOH/MeCN (50:50v/v)

C18 Luna column,25cm63.0mmID5 lm

Gradient Comparativestudyof5extraction methods(LLE andSPE)

UV, DAD;MS HPLCmethodandca-pillary electrophoresismethod. (HYTY, TY,oleuropein and ligstro-side aglycone, and dec-arboxymethyl oleuro-pein aglycone)

Bendini et al.[102]

65 min A: H2O + C H3COOH2%B: EtOH

PhenomenexLuna

(phenyl-hexyl)phase;25cm64.6mmID5 lm

Isocratic LLEwithmethanol/

water(80:20 v/v) Mon-tedoroet al.[20]

UV; MS(ESI in

negative ionmode)

Isolation of individual

polyphenolsto studysensory properties

Andreweset al.

[59]

Method UV: HPLC method of Brenes et al. [30] Gradient L LE with N,N-dimethyl-formamide (DMF)

DAD;fluores-cence

Phenolic composition ofcommercialvirgin oliveoils

Garcaet al.[155]

60 min A: H2O + C H3COOH0.2% (pH3.1)B: MeOH

Inertsil ODS-3;15 cm64.6mmID5 lm


Photodiode ar-ray(280nmand339 nm)

Totalphenol contentafterstorage period(se-coiridoid derivativesand3, 4 DHPEA-AC)

Morellet al.[156]


15/22


complementary techniques; in fact, in several recent stud-ies the isolation and characterization of VOO phenoliccompounds have been done using HPLC and GC simulta-neously [79, 127, 158].

In Figure 2, it is possible to see the typical HPLC-phenolicprofile of two extracts of VOO [159]. These extracts wereobtained from extra-VOO obtained from Peranzana andNostrana di Brisighella olivecvv. In both samples, tyrosol,


Table 3. Continued.

Timeofanalysis

Mobi lephases Stationaryphase Typeofelut ion Extract ionsystem Detect ionsys-tem

Obse rvati ons Ref erences


C18 Luna column,25cm63.0mmID5 lm


DAD; MS(ESIin negative ionmode)

Effectof olive ripeningdegreeon theoxidativestablility and organolep-ticpropertiesof olive oil

Rotondiet al.[119]

HPLC method of Rotondi et al. [119] Gradient LLE with methanol/

water from oliveoil. SLEfromolive fruits.

DAD; MS(ESI

in positive andnegative ionmode)

HPLC andCE analysis.

3 simplephenols,a se-coiridoid derivative and2 lignans

Bonoliet al.

[176]

60 min A: H2O + C H3COOH0.5%B: MeOH/MeCN (50:50v/v)


Gradient ComparativestudyofLLEand SPE (diol andC18-phase)

Photodiode ar-ray detector;MS, NMR.

Simple phenols, secoiri-doidsand lignans

Hrnciriket al.[74]

60 min A: H2O + C H3COOHB: MeOH


Gradient Methanolicextractionand extractionwhich in-cludedSPE cleaningstep.

DAD;DAD/ESI-MS/MS

Olivefruits. Simple phe-nols,secoiridois, flavo-noidsand anthocya-nins.

Vinha etal.[101]

Figure 2. HPLC chromatogram at 280 nm of the extracts from Peranzana and Nostrana di Brisighellacv. olive oils. Peak identifi-cation:1, hydroxytyrosol;2, tyrosol;3, vanillic acid; 4, deacetoxyoleuropein aglycone zone;5, (+)-pinoresinol;6, (+)-1-acetoxypi-noresinol; I.S, internal standard (3,4-dihydroxyphenylacetic acid). From reference [157]. Mobile phase A, water/formic acid (99.5/0.5v/v); mobile phase B, acetonitrile. Column: C18LunaTM column 5 lm, 25 cm63.0 mm ID (Phenomenex). (Other conditionsin [157].)


16/22


hydroxytyrosol, and vanillic acid were detected; more-over, the extract from Peranzana oil showed a high con-centration of deacetoxyoleuropein aglycone (DAOA),while the Nostrana di Brisighella sample showed two lig-nans; (+)-pinoresinol and (+)-1-acetoxypinoresinol.

Compound elution is typical of reversed-phase HPLC, i.e.

polar compounds (e.g., simple phenols and phenolicacids) elute first, followed by those of decreasing polarity.

5 Capillary electrophoresis: A useful technique for

the analysis of olive oils

Many works in the literature, as has been said before,determine the total phenolic amount in olive oils by spec-trophotometric analysis and characterize their phenolicpatterns by capillary gas chromatography (CGC) and,mainly, by reverse phase high performance liquid chroma-tography (RP-HPLC). Even if the characterization andquantification of phenolic compounds from olive oils have

been successfully carried out by HPLC, this analyticaltechnique needs accurate sample preparation and, gen-erally, is time-consuming due to the complexity of thematrix. Actually, in the worst case the whole HPLC anal-ysis of olive oil's phenols requires more than 90 min.

Therefore, the use of faster analytical techniques andscreening tools, allowing a rapid screening of phenoliccompounds of olive oils, is strongly recommended. CEcan represent a good compromise between analysis timeand satisfactory characterization for some classes of phe-nolic compounds in VOO.

CE has become an alternative or complementary to theHPLC technique. The speed, resolution, and simplicity ofCE, combined with low operating costs, make the tech-nique an attractive option for the development of improvedmethods of food analysis for the new millennium [160166]. CE offers the analyst a number of key advantages

for the analysis of the components of food. Indeed, inrecent years, it has proved to be a high-resolution tech-nique and has been applied successfully to the analysis ofphenolic compounds of a large variety of samples (honey,plant extracts, wine, beer, tea, fruits, vegetables, juices )requiring only small amounts of sample and buffer and a

short analysis time [167172]. Unfortunately, few dataare available on the phenolic content in products that arewaste from the olive oil industry (olive mill waste-water [173] and alperujo [174]) and in olive oil [23,98,102,175, 176] obtained directly with this technique.

This review is presented as a summary of the most impor-tant publications in which a study on the polyphenoliccompounds in olive oil is carried out. For this reason, weonly summarize the information of the last mentionedreferences (see Table 4 and Table 5).

All the gathered methods use simple CZE methodologiesbased on a borate run buffer at alkaline pH. In the litera-

ture, the most efficient operative mode to separate pheno-lic compounds has been found to be borate-based CZE,but borate-phosphate-based micellar electrokinetic chro-matography (MECK) methods with sodium dodecylsulfate(SDS) as micellar agent have been also used [177 179].

Between all these methods, differences can be foundregarding applied voltage, internal diameter of the capil-lary, time of injection (in all the cases was hydrodynamicinjection), effective length of capillary, and buffer concen-tration. However, as previously stated, both type of bufferand pH are practically the same in all these methods. Anincrease in pH values causes higher migration times; infact, higher pH values lead to a higher ionization state ofthe species and at the selected basic pH, the phenoliccompounds are negatively charged and would migratetowards the anode, i.e. away from the detector. Neverthe-less, due to the large electroosmotic flow (EOF) in the sys-


Table 4.Summary of optimized conditions of capillary electrophoresis methods where olive oil samples are analyzed. kd, wave-length of detection; V, voltage;T, temperature, ID, internal diameter of capillary; Lef, effective length of capillary; [Buffer]; bufferconcentration.

Instrumental variables Experimental variables

References d [nm] V[kV] T [8C] ID [lm] Lef [cm] tinj [s] Type of buffer [Buffer][mM]

pH

Bendini et al. [102] 200 27 30 50 40 3 s (0.5p.s.i) Sodium tetra-borate 45 9.6

Bonoli et al. [175] CZE method of Bendini et al. [102]

Bonoli et al. [176] CZE method of Bendini et al. [102]

Carrasco Pancorboet al. [23]

210 25 25 75 50 8 s (0.5p.s.i)

Sodium tetra-borate

25 9.6

Buiarelli et al. [98] 200 18 25 50 36 2 s (1.5p.s.i)

Sodium tetra-borate

40 9.2


17/22


tem, polyphenols are propelled together with the bulk

solution towards the cathode, but at a much lower rate.Therefore, phenols and polyphenols are less dissociatedand those with a bigger molecular mass are first detectedsince they are less able to migrate upstream [168]. More-over, the lower analyte velocities observed at higherpH values could be explained by the increase in ionicstrength of the running buffer, which leads to lower elec-troosmotic flow [180].

For all these reasons, the best compromise in terms ofresolution of the phenolic compounds and total analysistime was obtained, in all cases, at about a pH 9.5 buffer.

The differences in internal diameter of the capillary and

time of injection, as well as the extraction system, thequantity of olive oil used in the extraction protocol, and thevolume of solvent, MeOH/H2O (50:50 v/v), for redissol-ving the phenolic compounds extracted, cause the differ-ences in sensitivity between these methods.

In three of these papers [102, 175, 176] the aim of theauthors is the study of compounds of all the most repre-sentative phenolic compounds in extra virgin olive oil,such as simple phenols (hydroxytyrosol and tyrosol andsome phenolic acids), secoiridoids (oleuropein aglyconederivatives), and lignans ((+)-1-acetoxypinoresinol). Theobjective for the authors of the other two papers [23, 98]was the characterization of a specific family among the

phenolic compounds, the phenolic acids. Recent interestin phenolic acids stems from their potential protective role,through ingestion of fruit and vegetables (e.g., olive oil),against oxidative damage diseases (coronary heart dis-ease, stroke, and cancers) [181].

Figure 3 shows one of the mentioned examples wherethe aim of the authors is the study of all the phenolic frac-tion of olive oil [159]. The figure shows the CE-profile oftwo extracts from extra-virgin olive oils obtained from Per-

anzana and Nostrana di Brisighella olives cvv.(they arethe same samples as those shown in the HPLC graphics;so it is possible to compare the two techniques). In this


Table 5.Summary of extraction systems used and detected compounds in olive oil samples with the application of each method.HYTY, hydroxytyrosol; TY, tyrosol; DHPE, 2,3-dihydroxyphenylethanol; VA, vanillic acid; DAOA, deacetoxy oleuropein aglycone;Ac Pin, (+)-1-acetoxypinoresinol.

References Extraction system Initial quantity of oil! Finalquantity of solvent (MeOH/H2O(50:50 v/v))intheextrac-tion process

Detected compounds in oliveoil

Bendini et al. [102] LLE (Pirisi et al. [73]) 2 g!1 mL HYTY, TY, unidentified secoir-idoids compounds

Bonoli et al. [175] LLE (Pirisi et al. [73]) 2 g! 1 mL HYTY, TY, DHPE, unidentifiedoleuropein aglycone deriv-atives

Bonoli e t a l. [176] LLE(Pirisi et a l. [ 73], m odifiedby Rotondi et al.[119])

2 g! 0.5 mL HYTY, TY, VA, DAOA, Ac Pin

Carrasco Pancorbo et al. [23] LLE [23] 60 g! 0.5 mL 14 phenolic acids

Buiarelli et a l.[98] Combination o f L LE-SPE [98] 10 g! non s pecified 5 p henolic a cids

Figure 3. CZE electropherograms at 200 nm of the extractsfrom Peranzana and Nostrana di Brisighella cvv. olive oils.See Figure 2 for analyte identification numbers. Separationconditions: capillary, 47 cm650 lm; applied voltage, 27 kV;applied temperature, 308C; buffer, 45 mM Sodium borate(pH 9.60); hydrodynamic injection, 0.5 p.s.i. for 3 s. Fromreference [157].


18/22



Figure 4. CZE electropherogram of phenolic fraction extracted from real olive oil sample by LLE. (a) Arbequina, (b) Lechn ofSevilla, (c) Picual, (d) Hojiblanca, (e) Lechn of Granada, (f) Cornicabra, (g) refined olive oil and (h) mixture of refined and virginolive oil. From reference [51]. Separation conditions: capillary, 57 cm675lm; applied voltage, 25 kV; applied temperature,258C; buffer, 25 mM Sodium borate (pH 9.60); hydrodynamic injection, 0.5 p.s.i. for 8 s. Detection was performed at 210 nm.Peak identification numbers: 1,trans-cinnamic acid;2, 4-hydroxyphenylacetic acid; 3, sinapinic acid; 4, gentisic acid;5, (+)-taxi-folin; 6, ferulic acid; 7, o-coumaric acid; 8, p-coumaric acid; 9, vanillic acid; 10, caffeic acid; 11, 4-hydroxybenzoic acid; 12,dopac;13, gallic acid and14, protocatechuic acid.


19/22


case, the elution order, according to the different principleof separation, was different from the HPLC one. Tyrosoland hydroxytyrosol were recognized in both oils. Similarlyto that detected by HPLC, deacetoxy oleuropein aglycone(DAOA) by CZE was the highest peak of the Peranzanaextract. Furthermore, the CZE-DAOA peak was particu-

larly sharp and well resolved (at the baseline) from theother peaks, as well in the Nostrana di Brisighella extract.

This work reports a qualitative comparison betweenHPLC and CE separations of extra-VOO phenolic com-pounds. The main advantage of the CE separationreported was the short analysis time, with respect to theHPLC method. Actually, tyrosol, hydroxytyrosol, (+)-1-acetoxypinoresinol and DAOA were analyzed within10 min by CZE, while HPLC separation required up to75 min. A preliminary quantification of phenols realized bythe two techniques showed good agreement, confirmingthe correct assignment of the CZE peaks.

Figure 4 shows one of the examples mentioned where

the aim of the authors is the study of a specific family inthe phenolic fraction of olive oil [23]. A sensitive, rapid,efficient, and reliable method for the separation and deter-mination of phenolic acids by CZE was developed andapplied to 6 monovarietal extra-virgin olive oil samples,different refined olive oils, and commercial mixtures ofrefined and virgin olive oils in order to compare theamounts of phenolic acids. The differences in the phenolicacid profiles shown in the electropherograms of the differ-ent varieties of olive oil extracts are very clear. It is animportant factor to bear in mind, in order to compare theanalysis of the samples correctly, that the absorbancescales of the different electropherograms for each variety

of olive oil are different. When refined olive oil and mix-tures of refined and virgin olive oil were analyzed, theamounts of phenolic compounds obtained were smallerthan for the other types of olive oils.

Compounds such as trans-cinnamic acid, sinapinic acid,(+)-taxifolin, caffeic acid, and dopac, which only appearedin several olive oils, could be considered as potential mar-kers for geographical origin or olive fruit varieties in thefuture.

In this work [23] the potential of the CE technique with UVdetection for fast and sensitive simultaneous determina-tion of 14 compounds of the same family in extra-virgin

olive oils obtained from different varieties has beendemonstrated.

6 Conclusions

Polyphenols are significantly related to the quality of virginolive oil and their contribution to the oxidative stability ofthe oil is widely accepted. The qualitative and quantitativecomposition of VOO hydrophilic phenols is strongly

affected by the agronomic and technological conditions ofits production. For these reasons, the identification andquantification of the individual components of VOO are ofgreat interest. Many analytical procedures directedtowards the determination of the complete phenolic profilehave been proposed; in general, an analytical procedure

for the determination of individual phenolic compounds inVOO involves three basic steps: extraction from the oilsample, analytical separation, and quantification. Thevariety of extraction techniques, chromatographic condi-tions, and methods of quantification have contributed tothe differences in reported levels of virgin olive oil phenoliccompounds.

The use of capillary electromigration methods to analyzephenolic compounds of olive oil is nowadays increasing,although the usual procedure now encompasses a high-performance separation technique in combination withdiode array detection or mass spectrometry.

The high resolution, efficiency, and analysis speed pro-

vided by CE together with the minimum sample andreagents consumption have promoted the use of this tech-nique. Also the suitability of CE for coupling to differenttypes of detectors may make CE a powerful tool for thecharacterization of the phenolic fraction of olive oil.

References

[1] D. Grigg, GeoJournal. 2001, 53, 163 172.

[2] E. Helsing, Europ. J. Clin. Nutr. 1993, 47, S4S12.

[3] Commission Regulation 1513/2001, 23th July. Off. J. Eur.Communities. 2001,L201,47.

[4] E. Fedeli, Prog. Chem. Fats Other Lipids1977, 15,5774.

[5] J.B. Rossell,Frying: Improving quality. Woodhead Publish-ing Limited, CRC Press, Boca Raton. Boston. New YorkWashington, DC, 2001.

[6] E. Tiscornia, M. Forina, F. Evangelisti, Riv. Ital. SostanzeGrasse1982, 59, 519 556.

[7] F. Visioli, C. Galli,J. Agric. Food Chem. 1998, 46, 42924296.

[8] M. Servili, G. Montedoro,Eur. J. Lipid Sci. Technol. 2002,104, 602 613.

[9] D. Boskou, Olive oil: Chemistry and Technology. AOCSPress,Champaign,IL (USA), 1996. 52 83.

[10] F. Shahidi,Natural antioxidants: Chemistry, health effectsand applications. AOCS Press, Champaign, IL (USA), 1997,97149.

[11] J.B. Harborne, P.M. Dey, Methods in Plant Biochemistry.J.B. Harborne(Ed.). Academic Press, London 1989.

[12] E. Psomiadou, X. Konstantinos, K.G. Blekas, M.Z. Tsimi-dou, D. Boskou,Eur. J. Lipid Sci. Technol. 2003,105, 403408.

[13] C. Cantarelli, Riv. Ital. Sost. Grasse1961, 38,6972.

[14] G.F. Montedoro, C. Cantarelli,Riv. Ital. Sost. Grasse1969,46, 115 124.

[15] G.F. Montedoro, Sci. Technol. Aliment. 1972, 3, 177 186.

[16] A. Vzquez Roncero, Rev. Fr. Corps Gras.1978, 25, 2126.



20/22


[17] M. Solinas, A. Cichelli,Riv. Ital. Sostanze Grasse1981,58,159164.

[18] N. Cortesi, E. Fedeli,Riv. Ital. Sostanze Grasse1983, 60,341351.

[19] M. Solinas, Riv. Ital.Sostanze Grasse1987, 64, 255 262.

[20] G.F. Montedoro, M. Servili, M. Baldioli, E. Miniati,J. Agric.Food. Chem. 1992, 40, 1571 1576.

[21] M. Tsimidou, M. Lytridou, D. Boskou, A. Paooa-Lousi, F.Kotsifaki, C. Petrakis, Grasas Aceites1996, 47, 151 157.

[22] S. Mannino, M.S. Cosio, M. Bertuccioli, Ital. J. Food. Sci.1993, 4, 363 370.

[23] A. Carrasco Pancorbo, C. Cruces-Blanco, A. Segura Carre-tero, A. Fernndez Gutirrez, J. Agric. Food. Chem.2004,52, 6687 6693.

[24] M.J. Garrido Fernndez Dez, M. R. Adamos,Table olives.Chapman & Hall, London(UK), 1997, pp. 67 109.

[25] G.F. Montedoro, M. Servili, M. Baldioli, E. Miniati,J. Agric.Food. Chem. 1992, 40, 1577 1580.

[26] G.F. Montedoro, M. Servili, M. Baldioli, R. Selvaggini, E.Miniati, A. Macchioni, J. Agric. Food. Chem. 1993, 41,22282234.

[27] F. Angerosa, N. DAlessandro, P. Konstantinou, L. Di Gia-

cinto, J. Chromatogr. A 1996, 736, 195 203.[28] R.W. Owen, W. Mier, A. Giacosa, W.E. Hull, B. Spiegel-halder, H. Bartsch, Food Chem. Toxicol. 2000, 38, 647659.

[29] E. Perri, A. Raffaelli, G. Sindona, J. Agric. Food Chem.1999, 47, 4156 4160.

[30] M. Brenes, A. Garca, P. Garca, J.J. Ros, A. Garrido,J.Agric. FoodChem. 1999, 47, 3535 3540.

[31] R. Mateos, J.L. Espartero, M. Trujillo, J.J. Ros, M. Len-Camacho, F. Alcudia, A. Cert, J. Agric. Food Chem.2001,49, 2185 2192.

[32] A.D. Bianco, L. Muzzalupo, G. Romeo, M.L. Scarpati, A.Soriero, N. Uccella, FoodChem. 1998, 63, 461 464.

[33] P. Rovellini, N. Cortesi, E. Fedeli, Riv. Ital. Sostanze Grasse1997, 74, 273 279.

[34] A. Vzquez-Roncero, C. Janer, M.L. Janer,Grasas Aceites1976, 27, 185 191.

[35] R.W. Owen, W. Mier, A. Giacosa, W.E. Hull, B. Spiegel-halder, H. Bartsch, Clin.Chem. 2000, 46, 976 988.

[36] M. Brenes, F.J. Hidalgo, A. Garca, J.J. Ros, P. Garca, R.Zamora, A. Garrido,J. Am. Oil Chem. Soc.2000,77, 715 720.

[37] A. Bianco, F. Coccioli, M. Guiso, C. Marra,Food Chem.2001, 77, 405 411.

[38] R. Leenen, A.J.C. Roodenburg, M.N. Vissers, J.A.E.Schuurbiers, K.P.A.M. van Putte, S.A. Viseman, F.H.M.M.vande Put, J. Agric.Food Chem. 2002, 50, 1290 1297.

[39] M.N. Vissers, P.L. Zock, R. Leenen, A.J.C. Roodenburg,K.P.A.M. van Putte, M.B. Katan, Free Radical Res. 2001,35, 619 629.

[40] R. Briante, F. La Cara, M.P. Tonziello, F. Frebbraio, R.Nucci, J. Agric.FoodChem. 2001, 49, 3198 3203.

[41] A. Petroni, M. Blasevich, M. Salami, N. Napini, G.F. Monte-doro, C. Galli, Thromb. Res. 1995, 78, 151 160.

[42] F. Caponio, V. Alloggio, T. Gomes,Food Chem. 1999,64,203209.

[43] F. Caponio, T. Gomes, A. Pasqualone, Eur. Food Res.Technol. 2001, 212, 329 333.

[44] M. Tsimidou, Ital. J. Food Sci. 1998, 10, 99116.

[45] M. Baldioli, M. Servili, G. Perreti, G. F. Montedoro, J.Am. OilChem. Soc. 1996, 73, 1589 1593.

[46] J. Velasco, C. Dobarganes, Eur. J. Lipid Sci. Technol. 2002,104, 661 676.

[47] M. Soler, L. Chatenaud, C. La Vecchia, S. Franceschi, S.Negri, Eur. J. CancerPrev. 1998, 7, 461 464.

[48] S. Franceschi, A. Bavero, E. Conti,R. Salamini,R. Volpe,E.Negri, L. Barman, C. La Vecchia, Br. J. Cancer1999, 80,614620.

[49] E. Hodge, D.R. English, M.R.E. McCredie, G. Severi, P.Boyle, J.L. Hopper, G.G. Giles, Cancer Causes Control2004, 15,1122.

[50] F. Visioli, G. Bellomo, G.F. Montedoro, C. Galli, Athero-sclerosis1995, 117,2532.

[51] C. Manna, V. Galletti, P. Cucciolla, G.F. Montedoro, V. Zap-pia, J. Nutr. Biochem. 1999, 10, 159 165.

[52] R.W. Owen, A. Giacosa, W.E. Hull, R. Haubner, G. Wrtele,B. Spiegelhaldar, H. Bartsch, Lancet2000, 1, 107 112.

[53] R.W. Owen, A. Giacosa, W.E. Hull, R. Haubner, B. Spiegel-halder, H. Bartsch, Eur. J. Cancer2000, 36, 1235 1247.

[54] M.H. Gordon, F. Paiva-Martins, M. Almeida,J. Agric. Food.Chem. 2001, 49, 2480 2485.

[55] F. Gutirrez-Rosales, S. Perdiguero, R. Gutirrez, J.M.Olas, J. Am.Oil Chem. Soc. 1992, 69, 394 395.

[56] F. Gutirrez-Rosales, J.J. Ros, M.L. Gmez-Rey,J. Agric.FoodChem. 2003, 51, 6021 6025.

[57] G.F. Montedoro, M. Baldioli, M. Servili,Giornale Ital. Nutriz.Clin.Prev. 1992, 1,1932.

[58] D. Ryan, K. Robards, Analyst1998, 123, 31R 44R.[59] P. Andrewes, J.L.H.C. Busch, T. de Joode, A. Groenewe-

gen, H. Alexandre, J. Agric. Food Chem.2003,51, 14151420.

[60] E. Graciani-Costante, A. Vzquez-Roncero, Grasas Aceites1981, 32, 365 371.

[61] N. Uccella, in: A.H. Spanier, F. Shahidi, T.H. Parliment, C.J.Mussinan, C.T. Ho, E. Tratas Conti, (Eds.), in: Food Fla-vours and Chemistry: Advances of the New Millenium. TheRoyal Society of Chemistry Publishers, Cambridge, UK,2001, 253.

[62] A.K.Kiritsakis, J. Am. OilChem. Soc. 1998, 75,673681.

[63] J.M. Garca, K. Yousfi, R.Mateos, M.Olmo,A. Cert, J. Agric.Food. Chem. 2001, 49, 4231 4235.

[64] M.J. Tovar, M.J. Motilva, M. Paz Romero, J. Agric. Food.Chem. 2001, 49, 5502 5508.

[65] R. Mateos, A. Cert, M.C. Prez-Camino, J.M. Garca, J.Am.OilChem. Soc. 2004, 81,7175.

[66] L. Cinquanta, M. Esti, E. La Notte,J. Am. Oil Chem. Soc.1997, 74, 1259 1264.

[67] M. Esti, L. Cinquanta, E. La Notte, J. Agric. Food Chem.1998, 46,3235.

[68] M. Uceda, M. Hermoso, A. Garca-Ortiz, A. Jimnez, G.Beltrn, Acta Horticulturae1999, 474,659652.

[69] P. Capella, E. Fedeli, G. Bonaga, G. Lercker,Manuale deglioli e dei grasi. Tecniche Nuove, Milano (Italy), 1997.

[70] M. Servili, M. Baldioli, G.F. Montedoro, Acta Horticulturae

1994, 356, 331 336.[71] L. Di Giovacchino, S. Sestili, D. Di Vicenzo, Eur. J. Lipid Sci.

Technol. 2002, 104, 587 601.[72] A. Ranalli, F. Angerosa,J. Am. Oil Chem. Soc. 1996, 73,

417422.[73] F.M. Pirisi, P. Cabras, C. Falqui Cao, M. Migliorini, M. Mug-

gelli, J. Agric.Food Chem. 2000, 48, 1191 1196.

[74] K. Hrncirik,S. Fritsche, Eur. J. Lipid Sci. Technol. 2004, 106,540549.

[75] M. Brenes, L. Rejano, P. Garca, A.H. Snchez, A. Garrido,J. Agric.Food Chem. 1995, 43, 2702 2706.



21/22


[76] D. Tura, K. Robards, J. Chromatogr. A 2002, 975,7193.

[77] M. Solinas, A. Cichelli, Riv. Soc. Ital. Sci. Aliment. 1981, 10,159164.

[78] F. Evangelisti, P. Zunin, E. Tiscornia, R. Petacchi, G.Drava, S. Lanteri,J. Am. Oil Chem. Soc.1997,74, 10171023.

[79] M. Tasioula-Margari, O. Okogeri, J. Food. Sci. 2001, 66,530538.

[80] M. Tsimidou, G. Papadopoulos, D. Boskou, Food Chem.1992, 44,5360.

[81] A. Garca, M. Brenes, M.J. Moyano, J. Alba, P. Garca, A.Garrido, J. Food Eng. 2001, 48, 189 194.

[82] G. Beltrn, A. Jimnez, M.P. Aguilera, M. Uceda, GrasasAceites2000, 51, 320 324.

[83] A. Vzquez-Roncero, C. Janer, M.L. Janer, Grasas Aceites1973, 24, 350 357.

[84] G. Montedoro, C. Cantarelli, Riv. Ital. Sost. Grasse1969,46,312.

[85] N. Cortesi, M. Azzolini, P. Rovellini, E. Fedeli, Riv. Ital.Sost. Grasse1995, 7

Carrasco-Pancorbo 2005-review.pdf

Documents